3. Unhedgeable risk 1
Acknowledgements
This report was commissioned by the Investment Leaders Group (ILG) and the University of Cambridge Institute
for Sustainability Leadership (CISL). From three University of Cambridge research institutions: the Centre for
Risk Studies (CRS), the Centre for Climate Change Mitigation Research (4CMR) and the Judge Business School.
Authors include: Dr Andrew Coburn, Jennifer Copic, Prof. Douglas Crawford-Brown, Dr Aideen Foley, Dr Scott
Kelly, Dr Eugene Neduv, Prof. Daniel Ralph, Dr Farzad Saidi and Jaclyn Zhiyi Yeo.
The study design and editorial process were led by Dr Jake Reynolds (Director, Sustainable Economy, CISL), Rob
Lake (independent consultant) and Clarisse Simonek (Programme Manager, ILG, CISL). The study benefited
from valuable input from the members of the ILG as well as an external Advisory Panel consisting of Richard
Lewney (Managing Director, Cambridge Econometrics), Dr Mike Maran (Chief Science Officer, Catlin Group Ltd),
Tomi Nummela (Associate Director, Implementation Support, PRI), Michael Sheren (Senior Adviser, Bank of
England), Rick Stathers (Head of Responsible Investment, Schroders), and John Ward (Managing Director, Vivid
Economics). This report is based on original research in Kelly et al 2015 which can be found at: www.risk.jbs.
cam.ac.uk.
Contents
Executive Summary ............................................................................................................................................. 5
1 Introduction ....................................................................................................................................................7
1.1 Investor risk .................................................................................................................................................................7
1.2 Study aim and objectives............................................................................................................................................8
2 The sentiment scenarios ............................................................................................................................... 8
2.1 Two Degrees ................................................................................................................................................................9
2.2 Baseline ......................................................................................................................................................................10
2.3 No Mitigation .............................................................................................................................................................10
2.4 Potential scenario triggers .......................................................................................................................................11
3 Macroeconomic analysis ............................................................................................................................. 13
3.1 The Oxford Economics’ General Equilibrium Model (GEM) .................................................................................13
3.2 Variable descriptions ................................................................................................................................................14
3.3 Macroeconomic results ...........................................................................................................................................14
3.4 Long-term impacts ....................................................................................................................................................15
3.5 Economic conclusions .............................................................................................................................................17
4. Unhedgeable risk2
4 Investment portfolio analysis ..................................................................................................................17
4.1 Valuation fundamentals .........................................................................................................................................17
4.2 Passive investor assumption .................................................................................................................................17
4.3 Standardised investment portfolios .....................................................................................................................18
4.4 High Fixed Income portfolio ...................................................................................................................................18
4.5 Conservative portfolio ............................................................................................................................................19
4.6 Balanced portfolio ...................................................................................................................................................20
4.7 Aggressive portfolio ................................................................................................................................................21
4.8 Computation of returns ..........................................................................................................................................22
4.9 Portfolio returns ......................................................................................................................................................23
4.10 Impact on returns – by asset class and geography ............................................................................................ 24
4.11 Impact on returns – by industry sector ................................................................................................................ 25
4.12 Financial conclusions .............................................................................................................................................28
5 Conclusion ...................................................................................................................................................30
6 References ...................................................................................................................................................30
Appendix
Appendix A Literature review .......................................................................................................................... 33
Appendix B Climate risk .................................................................................................................................... 33
B.1 The climate science ................................................................................................................................................. 33
B.2 Looking ahead .......................................................................................................................................................... 34
B.3 Opportunities for stabilisation and climate change commitment .................................................................... 35
B.4 Climate change adaptation ..................................................................................................................................... 35
B.5 The economics of climate change ......................................................................................................................... 36
B.6 Climate change risks ............................................................................................................................................... 36
B.7 The financial and economic risks ........................................................................................................................... 37
Appendix C Modelling and methods ................................................................................................................ 40
C.1 Developing a framework for analysis .................................................................................................................... 40
C.2 Modelling timeframe ............................................................................................................................................... 41
C.3 Modelling process .................................................................................................................................................... 41
C.4 Description of scenario stress testing analysis .................................................................................................... 43
C.5 Complex risks and macroeconomic impacts ........................................................................................................ 43
C.6 Developing coherent scenarios ............................................................................................................................. 44
5. Unhedgeable risk 3
C.7 A structural modelling methodology .................................................................................................................... 44
C.8 Assumptions and limitations ..................................................................................................................................45
C.9 Macroeconomic assumptions .................................................................................................................................45
Appendix D Climate change impacts ............................................................................................................... 46
D.1 Background ...............................................................................................................................................................46
D.2 Zonal climate statistics ............................................................................................................................................46
D.3 Derivation of impacts and risks ............................................................................................................................. 47
Appendix E Defining the scenarios .................................................................................................................. 50
E.1 Defining a climate risk sentiment scenario ........................................................................................................... 50
E.2 Selection process ......................................................................................................................................................50
Appendix F Background on macroeconomic modelling ............................................................................... 53
List of figures
Figure 1: Estimated modelled economic loss in global output (GDP@Risk) ........................................................ 15
Figure 2: Comparison of the long-term GDP projections across the scenario variants ...................................... 16
Figure 3: Asset class composition and geographic market spread in the High Fixed Income Portfolio
structure ........................................................................................................................................................ 19
Figure 4: Detailed asset class breakdown of the High Fixed Income Portfolio structure .................................... 19
Figure 5: Asset class composition and geographic market spread in Conservative Portfolio ............................ 20
Figure 6: Detailed asset class breakdown of Conservative Portfolio structure .................................................... 20
Figure 7: Asset class composition and geographic market spread in Balanced Portfolio .................................... 21
Figure 8: Detailed asset class breakdown of Balanced Portfolio ............................................................................ 21
Figure 9: Asset class composition and geographic market spread in Aggressive Portfolio ................................. 22
Figure 10: Detailed asset class breakdown of Aggressive Portfolio ......................................................................... 22
Figure 11: Comparison of total portfolio returns across sentiment scenarios ...................................................... 23
Figure 12: Comparison of equity performance by geography in nominal per cent change across sentiment
scenarios....................................................................................................................................................... 24
Figure 13: Comparison of fixed income performance by geography in nominal per cent change across sentiment
scenarios ...................................................................................................................................................... 25
Figure 14: US market equity performance across industry sectors – No Mitigation scenario ............................. 27
Figure 15: China market equity performance across industry sectors – No Mitigation scenario ........................ 27
Figure 16: US market equity performance across industry sectors – Two Degrees scenario .............................. 28
Figure 17: China market equity performance across industry sectors – Two Degrees scenario .......................... 28
6. Unhedgeable risk4
Figure 18: Climate change country credit risk, figure from (Standard and Poor’s, 2014) ...................................... 39
Figure 19: Conceptual links between 4CMR and CRS in the co-developed structural modelling methodology
framework. ................................................................................................................................................... 41
Figure 20: Structural modelling methodology for developping coherent stress test scenario............................. 45
Figure 21: Regional mean temperature anomalies under RCP8.5 and RCP2.6, for the period 2071-2100 relative
to the pre-industrial period. ....................................................................................................................... 47
Figure 22: IPCC RCP scenarios (Source: Stocker, 2014) ............................................................................................. 50
Figure 23: 3D scenario matrix approach (Adapted from van Vuuren et al., 2013) ................................................ 52
List of tables
Table 1: Potential triggers that may lead to the development of each scenario .................................................. 12
Table 2: Summary of macroeconomic impacts across the sentiment scenarios modelled ................................ 15
Table 3: Long-term impact with respect to baseline ................................................................................................ 16
Table 4: Summary of the High Fixed Income portfolio asset allocation ................................................................ 18
Table 5: Summary of the Conservative portfolio asset allocation .......................................................................... 19
Table 6: Summary of the Balanced portfolio asset allocation ................................................................................. 20
Table 7: Summary of the Aggressive portfolio asset allocation .............................................................................. 21
Table 8: Worst performing industry sectors across both developed and emerging economies ........................ 26
Table 9: Best performing industry sectors across both developed and emerging economies .......................... 26
Table 10: Summary of portfolio performance measured by the five per cent VaR by structure and scenario,
nominal per cent ........................................................................................................................................... 29
Table 11: Summary of portfolio performance (long-term impact after five years) by structure and scenario,
nominal per cent ........................................................................................................................................... 29
Table 12: Summary of climate risk-related research papers and reports ............................................................... 33
Table 13: Projected change in global mean surface air temperature and global mean sea level rise for mid and
21st century relative to the reference period of 1986-2005 .................................................................... 34
Table 14: Remaining production capacity in the agriculture sector following any amount of temperature increase
relative to pre-industrial revolution, for each of the world regions considered here .......................... 49
Table 15: Summary damage ratio estimated from the climate model across defined sectors for 2080 – 2110:
Combined effects of temperature and SPEI ............................................................................................... 49
Table 16: The extended scenario matrix architecture (Source: CRS)........................................................................ 51
Table 17: Key input variables and their maximum shocks applied to the respective scenario
variants ........................................................................................................................................................... 53
7. Unhedgeable risk 5
Executive Summary
Commissioned by the University of Cambridge Institute for Sustainability Leadership (CISL) and the Investment
Leaders Group (ILG), this report looks at the economic and financial impacts of climate risk over the next
five years in order to identify opportunities for reducing climate-related investment risks through portfolio
construction and diversification across different asset classes, regions and portfolios.
While the most significant physical impacts of climate change will probably be seen in the second half of this
century, financial markets could be affected much sooner, driven by the projections of likely future impacts,
changing regulations and shifting market sentiment. This study employs a unique approach to address these
short-term implications of our long-term climate problem in relation to portfolio risk. The complex analysis
presented here is the fruit of a collaborative effort between three research entities within the University of
Cambridge, namely the Cambridge Centre for Risk Studies (CRS), the Cambridge Centre for Climate Change
Mitigation Research (4CMR) and the Cambridge Judge Business School.
Both regulators and financial markets react in light of new information about climate change, including major
events such as storms, floods and droughts; policy decisions; and the success and failure of companies. While
such changes are partly visible in the present, they are likely to accelerate as the physical impacts of climate
change become deeper and more regular. This will influence financial market behaviour gradually in the first
instance (led by the most informed investors) and then, potentially, in a more disorderly fashion as markets
seek to shed at-risk assets. Some of the economic losses incurred by investors in this transition can be avoided
or hedged through mere reallocation strategies, while others require system-level intervention in the form of
policy or regulatory action. While we cannot model the psycho-social dynamics of financial markets despite the
ready availability of risk information, we can model physical impacts and consequences for the macroeconomy
under different scenarios of climate change mitigation identified by the Intergovernmental Panel on Climate
Change (IPCC) climate science. In doing so, our study characterises in detail the multi-faceted impact of climate
change on different asset classes and geographies.
Short-term shifts in market sentiment induced by awareness of future, as yet unrealised, climate risks
could lead to economic shocks, causing substantial losses in financial portfolio value within timescales
that are relevant to all investors.
Factors, including climate change policy, technological change, asset stranding, weather events and longer-
term physical impacts may lead to financial tipping points for which investors are not presently prepared.
This research shows that changing asset allocations among various asset classes and regions, combined
with investing in sectors exhibiting low climate risk, can offset only half of the negative impacts on financial
portfolios brought about by climate change. Climate change thus entails “unhedgeable risk” for investment
portfolios.
While the response to action aimed at limiting warming below 2°C is shown to be negative in its short-term
economic and financial impacts, the benefits of early action lead to significantly higher economic growth
rates and returns over the long run, especially when compared to a worst-case scenario of inaction. The
present study shows that certain types of portfolio benefit more than others.
Even in the short run, the perception of climate change represents an aggregate risk driver that must
be taken into consideration when assessing the performance of asset portfolios. Our analysis provides
investors with a general guide to minimising their exposure to climate sentiment risk and has the potential
to stimulate a constructive dialogue between investors, governments and regulators to examine the
conditions necessary to build more resilient financial markets under unprecedented environmental
change.
8. Unhedgeable risk6
The results of this analysis show that, on a worst-case No Mitigation basis, 47 per cent of the negative impacts of
climate change across industry sectors can be hedged through industrial sector diversification and investment
in industries that exhibit few climate-related risks. Similarly, shifting from an aggressive equity portfolio to one
with a higher percentage of fixed-income assets makes it possible to hedge 49 per cent of the risk associated
with equities. However, these two “halves” are not cumulative, such that no strategy will offer more than 50
per cent coverage. This gives rise to the conjecture that, even in the short run, climate change will constitute an
aggregate risk system-wide action in order to mitigate its economy- wide effects.
We adopt an approach that – as far as we are aware – has not been applied to analysis of the financial
implications of climate change: stress testing representative portfolios using economic and market confidence
shocks derived from climate change sentiment scenarios. This study quantifies the potential financial impacts
of a shift in market sentiment driven by significant changes in investor and consumer beliefs about the future
effects of climate change, modelling the impact of three market sentiment scenarios on four portfolios with
different asset allocations.
The scenarios reflect differing beliefs about the likelihood of government action to limit warming to 2°C, as
recommended by the IPCC, the actual emission levels anticipated, as well as physical climate change impacts,
the probable stringency of regulation and levels of investment, including the types of technology likely to be
developed. The scenarios, aptly named Two Degrees, No Mitigation, and Baseline, were developed according to
well-recognised risk modelling techniques, drawing on the latest IPCC climate change projections and employing
analysis of historical market shocks that offer meaningful parallels to interpret and model parameters within
a climate risk framework.
The asset allocations employed are representative of pension funds (Conservative, Balanced and Aggressive)
and insurance companies (High Fixed Income). Portfolios were stress tested by applying shocks based on
different levels of carbon taxation, energy investment, green investment, energy and food prices, energy
demand, market confidence, bond market stress and housing prices. Impacts are modelled over five years at
the portfolio, asset class, sector and regional levels. This analysis enables the quantification of impacts for each
scenario across different asset classes, industrial sectors, countries and portfolio structures.
The macroeconomic analysis shows that the transition to a low-carbon economy carries increased economic
costs in the short term, but that longer term discounted benefits make a transition more than worthwhile.
The sheer scale of structural change required for the global economy to shift away from a future dominated
by fossil fuels towards a low-carbon economy requires tremendous investment in new capital infrastructure,
in research and development, and in new business models. This transition period lasting years will be costly
to the global economy. However, the alternative may well be worse: results from the macroeconomic analysis
show that the No Mitigation scenario triggers a global recession for three consecutive quarters, shrinking the
global economy by as much as 0.1 per cent each quarter.
By comparison, the Two Degrees scenario grows at just 0.3 per cent per quarter, whereas the Baseline model
offers the fastest near-term growth, reaching 0.7 per cent per quarter. Over a longer time horizon (2015-2050),
however, the Two Degrees scenario is shown to outperform the Baseline by 4.5 per cent with a discount rate of
3.5 per cent. While the degree of benefit varies by portfolio type, all portfolios experienced short-term losses
and long-term benefits in this scenario. Clearly, it is only after the learning process, technological progress
and construction of new infrastructure systems are complete that the positive benefits of the new low- carbon
economy begin to accrue. Again, this process contrasts with the No Mitigation scenario, where economic output
never recovers, but is supressed indefinitely below Baseline.
Turning to the impact of these horizons on market players, our results show that the High Fixed Income
9. Unhedgeable risk 7
portfolio carries the least risk from financial market disruption, but also experiences low performance across
all scenarios. In comparison, if no action is taken to limit warming to 2°C, a Conservative portfolio with a 40 per
cent weighting to equities (typical of a pension fund) could suffer permanent losses of more than 25 per cent
within five years after the shock is experienced.
Sectoral analysis shows that some hedging of climate risk is possible by targeting low risk equity investments
across different regions and sectors of the economy. Overall, emerging markets are the worst affected. Results
from the models suggest that approximately half of the impact on returns attributable to climate change can be
hedged through cross-industry and regional investment in low climate risk sectors. At the multi-asset portfolio
level, only half the negative impact of climate change on returns can be hedged by changing asset allocation.
Our analysis – alongside data tables generated for all our scenarios – provides investors with guidelines for
minimising their exposure while, at the same time, stimulating a dialogue that goes beyond mere reallocation
of resources to build a more sustainable capital market in an economy that is subject to environmental change.
1 Introduction
As awareness of climate-related risks grows and gains traction, prudent asset owners and asset managers are
beginning to question how global environmental trends – such as increasing pressure on agricultural land,
food security, soil degradation, local water stress, and extreme weather events – will affect the macroeconomic
performance of countries and sectors, and how this will play out in financial markets.
To address these questions, and on behalf of the Investment Leaders Group (ILG), the Cambridge Institute
for Sustainable Leadership (CISL) has commissioned this research from three Cambridge institutions: the
Cambridge Centre for Climate Change Mitigation Research (4CMR), The Cambridge Judge Business School and
the Centre for Risk Studies (CRS) therein. 4CMR has notable capacities in climate change research, whereas the
Centre for Risk Studies, located within the Cambridge Judge Business School, has considerable expertise in the
simulation of financial and economic impacts of various types of risk.
1.1 Investor risk
Climate change poses a major risk to the global economy affecting the wealth and prosperity of all nations
around the world. It will have major impacts on the availability of resources, the price of energy, the vulnerability
of infrastructure and the valuation of companies. This collaborative report studies how global trends arising
from the possible impacts of climate change will lead to a shift in market sentiment in the short term. The study
quantifies the economic impacts across regions, sectors and different asset classes, and models the estimated
change in value for different asset portfolios. In principle, the financial impacts resulting from different forms
of risk exposure can be hedged through strategic asset allocation and portfolio construction. However, some
portion of said risk exposure remains effectively ‘unhedgeable’ due to the inherent systematic risks associated
with different climate change scenarios. Avoiding systematic risks will require system-wide approaches such as
climate change mitigation and adaptation measures at the local, regional, national and global levels.
Appendix A presents a collection of articles that have been published over the past decade that range from
identifying climate change as an emerging risk, to addressing climate risks in the economy and portfolio
management (Mercer, 2015; Rogers et al., 2015; Stathers and Zavos, 2015; Kraemer and Negrila, 2014;
Committee on Climate Change, 2014; Mercer, 2014; Guyatt, 2011; Wellington and Sauer, 2005).
10. Unhedgeable risk8
1.2 Study aim and objectives
The most significant geophysical impacts of climate change will most likely be observed in the second half of
this century. As the climate continues to warm, global impacts will accumulate over time resulting in higher
long-term risks, particularly if global average temperatures rise above 3°C (Pachauri, 2014). Financial markets,
however, could show the impact of risk aggregation much sooner, as the effects of climate change will be
driven by the projections of likely future impacts, changing regulation and shifting market sentiment. Therefore,
investors should not be deterred from identifying and managing impending climate-based sentiment risks in
present-day financial markets based on long-term climate change projections. In fact, investors who act now
may benefit from first-mover advantage, or at the very least, minimise their exposure to such risks which could
evolve even more rapidly than anticipated (possible if climate science advances allow the timing of ‘tipping
points’ for climatic instability to be predicted).
This study models the effect of market behaviour and how financial investment decisions will be made under
different climate change sentiment scenarios. Anticipating how the market may respond to long- term climate
risks attempts to bridge the gap between the geophysical impacts of climate change over the longer term and
the potential effects that climate risk may have on the economic and financial markets today. Therefore the
aim is to identify and quantify the financial impacts that may arise from a shift in market sentiment driven by
significant changes in investor and consumer beliefs about the future effects of climate change. Given certain
conditions about future expectations of climate risk, market players will be prompted to change their behaviour
today to reduce their exposure across different asset classes.
This study intends to shed light on the vulnerability and resilience of different asset portfolios to climate change
related risks. With this information, investors will be able to hedge risk and invest in assets with lower potential
of being affected by climate change risk.
While this document is meant to focus on the results of our analysis, we provide further details and background
on our methodology in the Appendix. Appendix A provides a literature review. Appendix B introduces climate
change as an emerging risk to asset owners and managers. Appendix C describes the methods and stress test
scenario modelling approach that was adopted. Appendix D introduces the long- term climate change impacts
on different sectors and regions of the world. Appendix E discusses the selection process for our stress testing
scenarios, to which we turn next, and Appendix F provides further background information on the underlying
macroeconomic model.
2 The sentiment scenarios
We use a triplet of (i) selected radiative forcing levels, (ii) socio-economic pathways and (iii) climate policy
assumptions to develop three different sentiment scenarios, namely Two Degrees (Section 1.3), Baseline (Section
1.4) and No Mitigation (Section 1.5). The fundamental assumption of these scenarios is that public expectations
and economic sentiments provide a bridge between the future geophysical impacts from climate change and
the markets today. Scenarios do not model changes in the evolution and dynamics of policy implementation as
each scenario unfolds over time. Rather, we assume the markets behave in a way that is consistent with each
future scenario coming true. In sociology this is often referred to as the problem of reflexivity and occurs when
the observations or actions of observers in the social system affect the very system they are observing. We
remove this contradiction by only looking at the behaviour of the markets under the condition that the markets
believe a particular future scenario is going to come true. In this way the present day reactions of the market
are consistent with the belief of each future scenario unfolding, leading to the fulfilment of the scenario. For
example, in the No Mitigation scenario the reactions of the market are consistent with the belief that there will
11. Unhedgeable risk 9
be no mitigation and the impacts of climate change will be substantial. As a consequence, we do not account
for the fact that under the No Mitigation scenario there is a higher chance that governments will react more
strongly and will therefore avoid the worst effects of climate change.
2.1 Two Degrees
Two Degrees describes a world collectively making relatively good progress towards sustainability, with
sustained efforts to achieve future socio-economic development goals. In this analysis it is defined as being
similar to RCP2.6 and SSP1 from IPCC AR5. Resource intensity and dependence on fossil fuels are markedly
reduced. There is rapid technological development (including clean energy technologies and yield-enhancing
technologies for land), reduction of inequality both globally and within countries, and a high level of awareness
regarding environmental degradation. The world believes that global warming will not raise the average
temperature by more than 2°C above pre-industrial temperatures, but not without significant expense.
With this in mind, the economy gears up to make a transition away from fossil fuels and towards a low- carbon
economy. However, a shift in long-term investment decisions from several key technology and financial
institutions causes volatility and uncertainty in the financial markets leading to a short period of turmoil
and lower growth. The economy goes through an arduous period of divesting away from fossil fuels, where
nearly half of coal and oil assets are predicted to become stranded (HSBC, 2013). As the economy successfully
restructures toward renewables, investors slowly regain confidence and the market recovers over the medium
term.
Regulation: The level of global cooperation for mitigation is high, well-coordinated and early (within the next
five years from 2016 to 2020). For example, the effort of climate policies aimed at reducing GHG emissions is
reflected by the carbon tax imposed internationally on fossil fuel-dominant energy supply; a global carbon
budget is allocated and set at 20 per cent of the total underground carbon reserves (Carbon Tracker Initiative,
2013).1
Most major countries adopt the following carbon mitigation targets:
$1002
/tonne CO2
of carbon tax to reflect the strength of climate policies aimed at reducing greenhouse
gases (GHG) emissions
Carbon budgets set at 20 per cent on existing reserves
80 per cent more investments3
in low-carbon technologies
No further investment (or subsidies) for fossil fuel extractions
Direct impacts: Carbon taxes are implemented as an additional tax (i.e. not revenue neutral), thus they will be
passed on to businesses and consumers, thereby reducing purchasing power, productivity, investment, and
the economy’s total output (Congressional Budget Office, 2013). However, these carbon taxes may be used
to offset budget deficits or invest in research and development boosting long-term productivity and have an
overall positive effect on the economy in the long run (e.g. the Congress of the United States estimates a carbon
tax of $20/tonne CO2
would raise $1.2 trillion in revenue during its first decade).4
Rapid improvements in energy
efficiency, a decrease in the cost of renewables and the development of new agricultural technology leads to a
significant reduction in carbon intensities and higher yields from agriculture. The short term economic outlook
1 Unburnable Carbon 2013: Wasted capital and stranded assets (Online) http://www.lse.ac.uk/GranthamInstitute/wp-content/uploads/2014/02/PB-
unburnable-carbon-2013-wasted-capital-stranded-assets.pdf [Accessed: 14 Jan 2015]
2 Centre for Energy and Climate Economics (Online) Available: http://www.rff.org/centers/energy_and_climate_economics/Pages/Carbon_Tax_FAQs.aspx
[Accessed: 10 Feb 2015]
3 Financing a low-carbon future (Online) Available: http://static.newclimateeconomy.report/wp-content/uploads/2014/08/NCE_Chapter6_Finance.pdf
[Accessed: 11 Feb 2015]
4 Congressional Budget Office, 2013. Effects of a Carbon Tax on the Economy and the Environment. https://www.cbo.gov/sites/default/files/44223_Carbon_0.
pdf
12. Unhedgeable risk10
remains in a state of turmoil while the economy goes through a phase of readjustment and capital is reinvested
in a new energy system. This leads to a period of low short-term growth caused by high volatility, stranded
assets and uncertainty in many long-term investments. This rocky period does not persist for long, and the
long-term outlook for the economy remains positive.
Negative 3 per cent sentiment shocks across each of the major economies for one year
Negative sentiments taper back to zero and rise into the positive zone as markets regain confidence over
the five-year period.
2.2 Baseline
Baseline is a world where past trends continue (i.e. the business-as-usual BAU scenario), where there is no
significant change in the willingness of governments to step up actions on climate change. However, the worst
fears of climate change are also not expected to materialise and temperatures in 2100 are only expected to
reach between 2°C and 2.5°C. It is most similar to the IPCC’s RCP6.0 and SSP2. There is some progress towards
reducing resource and energy intensity (compared to historic rates), while the economy slowly decreases its
dependence on fossil fuel. Development of low-income countries proceeds unevenly, with some countries
making relatively good progress while others are left behind. In general, global population continues to rise,
especially in low-income countries. However, due to the lack of unified expectations regarding the future of
either regulation relating to GHG emissions or real economic activities, there is little hope that any significant
changes will occur to the existing economic conditions or climate policies over the short term (e.g. 2016 – 2020).
Regulations: Global climate policy actions are delayed beyond the modelling period, with only intermediate
success in reducing vulnerability to climate challenges.
No carbon or oil tax
World fossil fuel energy supply/production remains unchanged
Fossil fuel-dominant energy investments remain unchanged
No technological advances to renewable energy sources
Direct impacts: None in the near future but become increasingly significant beyond 2060.
2.3 No Mitigation
In the No Mitigation scenario, the world is oriented towards economic growth without any special consideration
for environmental challenges, rather the hope that pursuing self-interest will allow adaptive responses to
any climate change impacts as they arise. This is most similar to RCP8.0 and SSP5. In the absence of climate
policy, the preference for rapid conventional development leads to higher energy demand dominated by fossil
fuels, resulting in high GHG emissions. Investments in alternative renewable energy technologies are low but
economic development is relatively rapid, driven mainly by higher investment in human and knowledge capital.
13. Unhedgeable risk 11
The initial market volatility is high due to significant uncertainty around the future impacts of climate change.
But as the world progresses under conventional development, there is growing realisation that the majority
of wealth generated from strong economic performance was squandered on short term consumption. Thus,
market confidence as to the future performance of the economy is adjusted downward, initiating a widespread
down-grade in stock price valuations reflecting a future of low growth and economic output.
Regulation: Global mitigation, man-made drivers of climate change is low and delayed, with no coordinated
action in attributing pricing strategies to GHG emissions and land use change. In the absence of climate policies
within the five-year modelling period (i.e. 2016 – 2020):
No carbon or oil tax
50 per cent increase in world fixed investment for energy extraction5
Direct impacts: Fossil fuels remain the dominant source of energy, incentivising rapid technological progress
in large-scale energy and natural resource exploration and extraction, significantly driving up the price of fossil
fuel energy. As the markets recognise the high growth potential in fossil fuel based assets, investment in fossil
fuel companies increase; this capital is then spent on further exploration and extraction by these companies
(ultimately leading to the High climate change scenario). Climate change - induced environmental degradation
and water stress reduce agricultural yields across many parts of the world. The increased water stress, resource
constraints, and other environmental factors further strain production capabilities as well as regional social
cohesion.
10 per cent per year increase in global demand for carbon energy sources
10 – 20 per cent per year6
increase in world agriculture prices due to higher cost and lower land-use availability
Sentiment shocks of up to Neg. 5 – 8 per cent across the major economies
2.4 Potential scenario triggers
Many potential triggers may initiate a financial tipping point brought about by climate change and unravel in a
similar fashion to one of the scenarios described above. These triggers illustrate the different types of signals
that serve as leading indicators of the instantaneous sentiment shocks in each scenario; where an indicator
is nationally located, the trigger would be the emergence of a significant number of countries exhibiting that
indicator. In Table 1 we highlight five categories from which financial tipping points may result in one or more
triggers are: new scientific evidence and technology, new policy announcements, new legal developments,
increased social awareness, and economic factors.
5 Fossil Fuel Euphoria (Online) Available: http://www.tomdispatch.com/post/175760/tomgramper cent3A_michael_klare,_the_latest_news_in_fossil_fuel_addiction/
[Accessed 11 Feb 2015]
6 FAO Food price index (Online) Available: http://www.fao.org/worldfoodsituation/foodpricesindex/en/ [Accessed 11 Feb 2015]
14. Unhedgeable risk12
Table 1: Potential triggers that may lead to the development of each scenario
Trigger Two Degrees No Mitigation
News cientific
evidence and
Technology
• New technological breakthrough in
low-carbon technology
(e.g. fusion, solar)
• Increased accuracy in the monitoring
and measurement of emissions for
attribution
• New scientific evidence on the
unstoppable and runaway effects
of climate change
• Thermohaline circulation shuts
down
• Permafrost melts releasing vast
quantities of methane into the
atmosphere
• Greenland and Antarctica ice sheet
begins to melt
• Glaciers begins to disappear
New policy
announcements
• Announcement of global agreement
to limit GHG with a tax or a cap
• Election of new political party that
pushes climate change mitigation
• Forced nationalisation of selected
state assets
• Commitment to stop the implicit
subsidy of fossil fuels
• Chaos and breakdown in global
discussion on GHG policy
• Continued subsidy and government
action to open new oil fields
• Rollback on the price of carbon
from all major economies (e.g.,
China, Europe, USA)
New legal
developments
• Introduction of new case law on the
legality of emitting CO 2 emissions
based on existing law
• Increase in the number of lawsuits
and liabilities placed against
companies that emit CO 2 or with
disregard for the environment
• Climate change mitigation legal
challenges defeated in court
Increased social
awareness
• Increasing social awareness on the
risk of GHG emissions and
increasing reputational risk for
• Increasing social awareness of
changing growing seasons and
lower agricultural yields
companies that emit GHG
• Increased social awareness and
pressure from shareholders,
employees and activists to reduce
emissions
• Increase mechani sation and
carbon-intensity on farmlands for
fear of failed crops
Economic factors • Achievement of price parity between
renewable technology and fossil
fuels
• Stranded fossil fuel assets
• Persistent low fossil fuel prices
• Clean technology bubble collapse
15. Unhedgeable risk 13
3 Macroeconomic analysis
The effects of human activity over the next two decades will either put the earth on a path to limiting an increase
in temperature below 2°C compared to pre-industrial era temperatures, or commit the planet to temperature
increases above 4°C or more. While the most likely scenario will be somewhere between these two extremes, it
is nonetheless important to conduct ‘what if’ analysis to understand the implications for what these scenarios
might indicate for the economy. Cost-benefit analysis is the most common approach used within the literature
for estimating the discounted future costs of climate change, and attempts to estimate all potential future
costs and benefits across different climate change scenarios before discounting these estimates into present
day dollars. These estimates are often used to calculate the marginal cost of emitting a tonne of CO2
. In this
analysis we do not conduct a cost-benefit analysis, nor do we use an Integrated Assessment Model (IAM) to
estimate the costs and benefits of different climate change scenarios. Instead, this report draws on several
other studies (Ackerman and Stanton, 2006; The Economist, 2015; 2008) to inform the development of the
sentiment scenarios and for conducting the macroeconomic analysis. As part of the long-term economic
impacts of each sentiment scenario, a net present value (NPV) calculation is completed to compare the long-
term impact on global GDP with respect to the Baseline scenario. At a discount rate of 6 per cent and for the
period 2015-2050 the Two Degrees scenario has an economic benefit over Baseline of 3.2 per cent, while the
No Mitigation scenario has a long-term cost to the economy of approximately 14 per cent.
We use sentiment indicators as surrogates for awareness or perception of the economic impacts of climate
risk by the public, policymakers and investors alike, in order to drive the macroeconomic analysis. The
ambiguous concept of economic sentiment is usually neglected by mainstream macroeconomics because it
is not a variable easily observable or quantifiable (van Aarle and Kappler, 2012), and often dismissed as a
psychological or subjective component of beliefs (Arias, 2014). Nevertheless, the effect of sentiment shifts on
the macroeconomy can be significant. In this analysis we attempt to capture market sentiment using plausible
stress test scenarios. A significant part of the sluggish recovery following a downturn can also be attributed to
the pessimistic view held by markets. Negative or risk averse actors play a significant role during recessions, e.g.
during the hyperinflation period of the 70’s and beginning of the 80’s. The pessimistic wave of weak confidence
reinforces the downturn by further driving growth rates downwards, even though economic fundamentals
may have already recovered (Arias, 2014).
3.1 The Oxford Economics’ General Equilibrium Model (GEM)
We use the Oxford Economics’ GEM,7
a quarterly-linked international econometric model, to examine how the
global economy reacts to shocks of various types. It is the most widely used international macroeconomic model
with clients including the IMF and World Bank. The model contains a detailed database with historical values
of many economic variables and equations that describe the systemic interactions among, the most important
economies of the world. Forecasts are updated monthly for the five-year, 10-year, and 25-year projections.
The Oxford Economics’ GEM is best described as an eclectic model, adopting Keynesian principles in the short
term and a monetarist viewpoint in the long term. In the short term, output is determined by the demand side
of the economy; and in the long-term, output and employment are determined by supply side factors. The
Cobb-Douglas production function links the economy’s capacity (potential output) to the labour supply, capital
stock and total factor productivity. Monetary policy is endogenised through the Taylor rule, when central banks
amend nominal interest rates in response to changes in inflation. Relative productivity and net foreign assets
determine exchange rates, and trade is the weighted average of the growth in total imports of goods (excluding
oil) of all remaining countries. Country competitiveness is determined from unit labour cost.
7 The Oxford Economics’ Global Economic Model (GEM) is a rigorous quantitative modelling tool for forecasting and scenario analyses. http://www.oxfordeconomics.com
16. Unhedgeable risk14
3.2 Variable descriptions
Using the Oxford Economics GEM, three independent scenarios simulating market sentiment shifts are
modelled: Two Degrees, No Mitigation and the Baseline. The Baseline sentiment scenario is the projection
trajectory updated regularly by Oxford Economics; it acts as the control in this study and represents the
reference economic projection for comparing Two Degrees and No Mitigation.
Table 17 in Appendix F shows the shocks applied to the eight macroeconomic variables, which are existing
parameters across countries and regions contained as part of the Oxford Economics’ GEM, in terms of
magnitude of the shock and the extent of spatial impact across each scenario.
Some countries are illustrated as having larger macroeconomic shocks applied as compared to others because
they are subjected to higher impacts due to varying exposure levels for a given mean global temperature
change. Moreover, developed countries have greater capital stocks and infrastructure to better withstand
temperature rises, thereby increasing their resiliencies and correspondingly reducing the magnitude of shocks
applied. While most macroeconomic variables are shocked across the five-year modelling period to simulate
the shift in consumer behaviour, consumption, and policy, higher volatility parameters such as confidence
and house price indices are shocked for one year. Thus, the model simulates both longer-term behavioural
changes, policy impacts and the instantaneous effects of rapid shifts in market behaviour.
3.3 Macroeconomic results
In this section, we present the macroeconomic outputs in the short term, 2015-2020, based on the GDP@Risk
calculations for each of the three sentiment scenarios. A simple long-term macroeconomic analysis, from 2015-
2050, is given in Section 1.10 followed by a comparison between the short and long- term outcomes in Section
1.11.
Macroeconomic and financial market results were derived using the Oxford Economic Model driven by a set of
exogenous macroeconomic input shocks. These input shocks represent changes in commodity prices, shifts in
climate policy and shifts in market sentiment from each of the scenario narratives. The sensitivity of our results
to the market confidence parameter is explored further in Box 1 below.
Table 2 summarises the key macroeconomic impacts of climate risks for each of the sentiment scenarios:
growth rates and GDP figures. By definition, the technical indicator of a recession is two consecutive quarters of
negative economic growth commonly measured by GDP. Table 2 illustrates a global recession occurs for up to
three quarters in the No Mitigation scenario, where the global economy shrinks by up to 0.1 per cent (Q-on-Q)
compared to the baseline quarterly growth projection of 0.7 per cent.
The main macroeconomic output modelled from GEM is a year-on-year projection of the global economy. The
impacts of each sentiment scenario are compared with the baseline projection in which no crisis occurs. The
difference in economic output over the five-year period between the baseline and each sentiment scenario
represents the GDP@Risk.
The primary figure representing the impact of this catastrophe is the GDP@Risk metric, which represents the
total difference in GDP between the baseline and the scenario-specified projections. The total GDP loss over
five years, beginning in the first quarter of 2016 during which the shock of climate risk is applied and then
sustained through to the last quarter of 2020, defines the GDP@Risk. This is further expressed as a percentage
of the total GDP. Table 2 also provides the GDP losses for each scenario globally and across selected countries,
both as the potential total lost economic output over five years, and and as a percentage of the total economic
output from 2016 to 2020.
17. Unhedgeable risk 15
Figure 1: Modelled economic output between 2015 and 2020 for each sentiment scenario
Figure 2 Comparison of the long -term GDP projections for each sentiment scenario
65
70
75
80
85
90
2013 2014 2015 2016 2017 2018 2019 2020
GlobalGDP(US$Tn)
Baseline
Two Degrees
No Mitigation
0
50
100
150
200
250
2015 2020 2025 2030 2035 2040 2045 2050
GlobalGDP(US$Tn)
Two Degrees
Baseline
No Mitigation
Table 2: Summary of macroeconomic impacts across the sentiment scenarios modelled
3.4 Long-term impacts
The long-term economic impacts for each sentiment scenario were also analysed based on the underlying
assumption that the sentiment shift in each case is consistent with the subsequent physical change in climatic
conditions. The starting point for estimating long-term impacts includes running the GEM for a further five
years to 2025. This allows the GEM, which is a general equilibrium model, to re-equilibrate to a new long-term
growth trajectory based on the new economic conditions brought about in the initial five years.
Figure 1: Modelled economic output between 2015 and 2020 for each sentiment scenario
Macroeconomic Impact
Sentiment Scenarios
Baseline Two Degrees No Mitigation
Global variables
Min. quarterly growth rate
(Global Recession Severity)
0.7 % 0.3 % -0.1%
Global Recession Duration N/A N/A 3 Qtrs
Economic output
5-yr GDP
(US$ Tn)
GDP@Risk
(%)
Global 407.3 8.9 2.2% 19.1 4.7%
United States 91.4 3.0 3.2% 7.6 8.4%
United Kingdom 14.3 0.3 1.7% 0.8 5.8%
Germany 19.3 0.3 1.5% 0.7 3.8%
Japan 29.5 0.1 0.4% 0.6 2.3%
China 51.1 2.4 4.7% 4.6 9.0%
Brazil 12.3 0.4 3.5% 0.8 6.3%
GDP@Risk
(US$ Tn)
GDP@Risk
(US$ Tn)
GDP@Risk
(%)
18. Unhedgeable risk16
Figure 1: Modelled economic output between 2015 and 2020 for each sentiment scenario
Figure 2 Comparison of the long -term GDP projections for each sentiment scenario
65
70
75
80
85
90
2013 2014 2015 2016 2017 2018 2019 2020
GlobalGDP(US$Tn)
Baseline
Two Degrees
No Mitigation
0
50
100
150
200
250
2015 2020 2025 2030 2035 2040 2045 2050
GlobalGDP(US$Tn)
Two Degrees
Baseline
No Mitigation
From 2025 to 2050 we assume the annual growth rate of the global economy remains stable at year 10 (2025)
annual growth rates. This compounding economic growth rate is then used to estimate annual global GDP
levels to 2050. The further into the future economic output is projected, the greater is the uncertainty of
those estimates. Therefore, maintaining economic growth at the best long-term estimate is a good first order
approximation absent any additional information.
The long-term annual growth rates estimated from year ten are 3.5 per cent, 2.9 per cent and 2.0 per cent
for the Two Degrees, Baseline and No Mitigation sentiment scenarios, respectively. Obviously higher annual
growth drives higher long-term cumulative outcomes, which we give details next.
The results of this analysis are shown in Figure 2. Over the longer time horizon from 2015 to 2050, the divergence
between the different sentiment scenarios becomes very pronounced. In fact the initial sentiment shock to the
global economy cumulated over the first five years is negligible relative to the longer term cumulative impacts
of climate change. We show that over the longer term, a rise in average global temperatures of not more than
2°C is beneficial to the global economy due to a higher annual economic growth rate than the Baseline No
Mitigation scenarios. The long-term cumulative costs and benefits for each scenario over the 35-year period
between 2015 and 2050 are shown in Table 3. When compared to the Baseline scenario using a discount rate of
3.5 per cent, the Two Degrees scenario is shown to have a 4.5 per cent beneficial global impact on cumulative
output, while the No Mitigation scenario would have a negative impact of 16 per cent.
Figure 2: Comparison of the long-term GDP projections for each sentiment scenario
Table 3: Long-term impact with respect to baseline
Long -term Impact of Scenarios with Respect to Baseline
Scenario No Discount Rate 3.5% Discount Rate 6% Discount Rate
Two Degrees 6.5% 4.5% 3.2%
No Mitigation -19% -16% -14%
19. Unhedgeable risk 17
3.5 Economic conclusions
Although the physical effects of climate change will have limited impacts on the economy over the next five to
10 years, the effects of climate policy and market sentiment may cause significant economic disruption. The
short-term economic impacts of both No Mitigation and Two Degrees result in negative consequences for the
global economy. The No Mitigation scenario causes a global recession for the first three quarters of the analysis
period, while the Two Degrees scenario does not cause a global recession but does cut economic growth in half
when compared to baseline.
In the No Mitigation scenario the economy suffers an economic loss that continues indefinitely representing
and ongoing loss to global economic output, while in the Two Degrees scenario the economy performs worse
than Baseline for the first eight to twelve years, but eventually recovers and grows much faster than Baseline.
Using a discount rate of 3.5 per cent over 35 years, the Two Degrees scenario outperforms the Baseline by 4.5
per cent, while the No Mitigation scenario against underpreforms the Baseline by 16 per cent.
4 Investment portfolio analysis
The macroeconomic effects of the climate impact sentiment scenarios will also inevitably impact financial
markets. This section considers the market impact of the scenarios and corresponding consequences for
investors in the capital markets.
The performance of bonds, equities and alternatives in different markets is estimated from the macroeconomic
modelling, and compared with the performance at the start of the modelling period.
4.1 Valuation fundamentals
The goal is to estimate how the fundamentals of asset values are likely to change as a result of various market
conditions. This analysis is not a prediction of daily market behaviour and does not take into account the wide
variations and volatility that can occur in asset values due to trading fluctuations and the mechanisms of the
market.
4.2 Passive investor assumption
Fundamentally, the analysis assumes a passive and “traditional” financial portfolio investment strategy, in which
the proportions of different types of assets are fixed in advance, and are held constant via real-time rebalancing
throughout a given period. This assumption is knowingly unrealistic, as an asset manager is expected to react to
changing market conditions in order to rebalance risk across sectors and geographies within an asset class, and
asset owners across asset classes, yet it is a useful exercise to consider what might happen to a fixed portfolio.
It also provides a necessary benchmark representation against which active fund managers can compare
the performance of dynamic strategies as well as informing what drives the behaviour of a fixed portfolio at
different times. This draws investors’ attention to whether improved portfolio management processes are
required and gives greater insights into designing the optimal investment strategy.
20. Unhedgeable risk18
4.3 Standardised investment portfolios
This study considers four typical high quality investment portfolios designed in consultation with the advisory
panel and representatives from the financial services industry and insurance. They are fictional representative
portfolios that mimic features observed in the investment strategies of insurance companies (High-Fixed
Income Portfolio) and pension funds (Aggressive, Balanced and Conservative). For example, the Conservative
Portfolio structure has 59 per cent of investments in sovereign and corporate bonds, of which 95 per cent are
rated A or higher (investment grade), 40 per cent in equity markets and commodities make up the remaining 1
per cent of the portfolio structure.
The portfolio structures cover several asset classes and are geographically diverse. Investments spread across
the developed markets of the US, UK, Germany and Japan, as well as the emerging markets of Brazil and China.
The 40 per cent allocated to equity investments correspond to investments in stock indices. The Wilshire 5000
Index (W5000), FTSE100 (FTSE), DAX (DAX) and Nikkei (N225) are used to represent equity investments in the
developed markets, while the Boverspa (BVSP) and SSE Composite Index (SSE) represent emerging markets.
The maturity for long-term fixed-income bonds is assumed to be 10 years, while that of the short-term bonds
either three months or two years, limited by the macroeconomic model outputs for each country.
4.4 High Fixed Income portfolio
Details of the High Fixed Income Portfolio are shown in Table 4, Figure 3 and Figure 4.
Table 4: Summary of the High Fixed Income portfolio asset allocation
Asset Class USA UK Germany Japan Brazil China World Total
Government 2 yr 8.0% 6.0% 4.5% 2.5% 1.5% 1.5% 0.0% 24.0%
Government 10 yr 7.0% 7.0% 5.5% 1.5% 1.0% 1.0% 0.0% 23.0%
Corporate 2 yr 3.0% 3.0% 3.0% 2.0% 1.5% 1.5% 0.0% 14.0%
Corporate 10 yr 6.0% 7.0% 3.0% 2.5% 1.5% 2.5% 0.0% 22.5%
RMBS 2 yr 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
RMBS 10 yr 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Equities 3.0% 2.5% 3.0% 1.0% 1.0% 1.0% 0.0% 11.5%
Cash 4.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 4.0%
Commodities 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 1.0% 1.0%
Total 31.0% 25.5% 19.0% 9.5% 6.5% 7.5% 1.0% 100.0%
21. Unhedgeable risk 19
Fixed
Income
84%
Equity
12%
Cash
4%
Commodities
1%
31%
26%
19%
10%
7%
8%
Government 2 yr,
24.0%
Government 10 yr,
23.0%
Corporate 2 yr,
14.0%
Corporate 10 yr,
22.5%
Equities, 11.5%
Cash, 4.0%
Figure 3: Asset class composition and geographic market spread in the High Fixed Income Portfolio structure
Figure 4: Idem detailed asset class breakdown of the High Fixed Income Portfolio structure
4.5 Conservative portfolio
Details of the Conservative Portfolio are shown in Table 5, Figure 5 and Figure 6.
Table 5: Summary of the Conservative portfolio asset allocation
Asset Class USA UK Germany Japan Brazil China World Total
Government 2 yr 4.0% 2.0% 3.0% 0.0% 1.0% 1.0% 0.0% 11.0%
Government 10 yr 3.0% 2.0% 3.0% 1.0% 1.0% 1.0% 0.0% 11.0%
Corporate 2 yr 5.0% 4.0% 5.0% 2.0% 1.0% 1.0% 0.0% 18.0%
Corporate 10 yr 5.0% 4.0% 5.0% 3.0% 1.0% 1.0% 0.0% 19.0%
RMBS 2 yr 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
RMBS 10 yr 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Equities 19.0% 8.0% 8.0% 5.0% 0.0% 0.0% 0.0% 40.0%
Cash 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Commodities 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 1.0% 1.0%
Total 36.0% 20.0% 24.0% 11.0% 4.0% 4.0% 1.0% 100.0%
22. Unhedgeable risk20
Fixed
Income
59%
Equity
40%
Commodities 1%
36%
20%
24%
11%
4%
4%
Government 2 yr,
11.0%
Government 10 yr,
11.0%
Corporate 2 yr,
18.0%
Corporate 10 yr,
19.0%
Equities, 40.0%
Figure 5: Asset class composition and geographic market spread in Conservative Portfolio
Figure 6: Detailed asset class breakdown of Conservative Portfolio structure
4.6 Balanced portfolio
Details of the Balanced Portfolio are shown in Table 6, Figure 7 and Figure 8.
Table 6: Summary of the Balanced portfolio asset allocation
18
Figure 5: Asset class composition and geographic market spread in Conservative Portfolio
Figure 6: Detailed asset class breakdown of Conservative Portfolio structure
1.17 Balanced portfolio
Details of the Balanced Portfolio are shown in Table 6, Figure 7 and Figure 8.
Table 6: Summary of the Balanced portfolio asset allocation
Asset Class USA UK Germany Japan Brazil China World Total
Government 2 yr 3.0% 1.0% 3.0% 1.0% 0.5% 0.5% 0.0% 9.0%
Government 10 yr 3.0% 1.0% 3.0% 1.0% 0.5% 0.5% 0.0% 9.0%
Corporate 2 yr 4.0% 3.0% 5.0% 2.0% 0.5% 0.5% 0.0% 15.0%
Corporate 10 yr 5.0% 2.0% 5.0% 1.0% 0.5% 0.5% 0.0% 14.0%
RMBS 2 yr 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
RMBS 10 yr 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Equities 23.0% 9.0% 9.0% 4.0% 2.0% 3.0% 0.0% 50.0%
Cash 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Commodities 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 3.0% 3.0%
Total 38.0% 16.0% 25.0% 9.0% 4.0% 5.0% 3.0% 100.0%
Fixed
Income
59%
Equity
40%
Commodities 1%
36%
20%
24%
11%
4%
4%
Government 2 yr,
11.0%
Government 10 yr,
11.0%
Corporate 2 yr,
18.0%
Corporate 10 yr,
19.0%
Equities, 40.0%
23. Unhedgeable risk 21
Fixed
Income
47%
Equity
50%
Commodities
3%
38%
16%
25%
9%
4%
5%
Government 2 yr,
9.0%
Government 10 yr,
9.0%
Corporate 2 yr, 15.0%
Corporate 10 yr,
14.0%
Equities, 50.0%
Figure 7: Asset class composition and geographic market spread in Balanced Portfolio
Figure 8: Detailed asset class breakdown of Balanced Portfolio
4.7 Aggressive portfolio
Details of the Aggressive Portfolio are shown in Table 7, Figure 9 and Figure 10.
Table 7: Summary of the Aggressive portfolio asset allocation
19
Figure 7: Asset class composition and geographic market spread in Balanced Portfolio
Figure 8: Detailed asset class breakdown of Balanced Portfolio
1.18 Aggressive portfolio
Details of the Aggressive Portfolio are shown in Table 7, Figure 9 and Figure 10.
Table 7: Summary of the Aggressive portfolio asset allocation
Asset Class USA UK Germany Japan Brazil China World Total
Government 2 yr 2.0% 1.0% 1.0% 0.5% 0.5% 0.5% 0.0% 5.5%
Government 10 yr 2.0% 1.0% 1.0% 0.5% 0.5% 0.5% 0.0% 5.5%
Corporate 2 yr 4.0% 2.5% 2.0% 1.5% 1.0% 1.0% 0.0% 12.0%
Corporate 10 yr 4.0% 2.5% 2.0% 1.5% 1.0% 1.0% 0.0% 12.0%
RMBS 2 yr 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
RMBS 10 yr 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Equities 26.0% 11.0% 10.0% 5.0% 4.0% 4.0% 0.0% 60.0%
Cash 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Commodities 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 5.0% 5.0%
Fixed
Income
47%
Equity
50%
Commodities
3%
38%
16%
25%
9%
4%
5%
Government 2 yr,
9.0%
Government 10 yr,
9.0%
Corporate 2 yr, 15.0%
Corporate 10 yr,
14.0%
Equities, 50.0%
24. Unhedgeable risk22
Fixed
Income
35%
Equity
60%
Commodities
5%
38%
18%
16%
9%
7%
7%
Government 2 yr, 5.5%
Government 10 yr,
5.5%
Corporate 2 yr, 12.0%
Corporate 10 yr, 12.0%
Equities, 60.0%
Figure 9: Asset class composition and geographic market spread in Aggressive Portfolio
Figure 10: Detailed asset class breakdown of Aggressive Portfolio
4.8 Computation of returns
Portfolio returns are estimated using the following method.
Market price changes or Mark to Market (MtM) are calculated for all government bonds using equation (1) and
for corporate bonds and RMBS using equation (2):
1 ΔMtMGov,t
= (Db
) (-ΔI/100) 
2 ΔMtMCorp,t
= (Db
) (-ΔI/100) + (SDb
) (-ΔCS/100)
Here, Db
is the bond duration, for which we assumed the following values: Db
= 7 for ten-year bonds, Db
= 1.8
for two-year bonds, and Db
= 0.4 for three-month bonds. SDb
represents the spread duration. The change in
interest rates, ΔI on government and corporate bonds and the change in credit spreads, ΔCS, are taken from
the output of the macroeconomic analysis discussed in the previous chapter.
25. Unhedgeable risk 23
21
chapter.
Government bond yields are estimated using a representative quarterly yield while corporate and RMBS
yields are estimated using a representative quarterly yield and the period average d credit spread.
Equities market prices are calculated using the change in equity value from the macroeconomic
modelling. The equity dividends are es timated using a representative quarterly yield. Exchange rate
effects are taken into account to ensure all reported portfolio returns are illustrated in US dollars.
1.20 Portfolio returns
Figure 11: Comparison of total portfolio returns across sentiment scenarios
Figure 11 shows the scenario impacts by variant across all portfolio structures. We compared the impacts
of the scenarios in terms of total portfolio nominal returns . Across all portfolio structures and scenarios,
Government bond yields are estimated using a representative quarterly yield while corporate and RMBS yields
are estimated using a representative quarterly yield and the period averaged credit spread.
Market prices for equities are calculated using the change in equity value from the macroeconomic modelling.
Dividends are estimated using a representative quarterly yield. Exchange rate effects are taken into account to
ensure all reported portfolio returns are illustrated in US dollars.
4.9 Portfolio returns
Figure 11: Comparison of total portfolio returns across sentiment scenarios
Figure 11 shows the scenario impacts by variant across all portfolio structures. We compared the impacts of
the scenarios in terms of total portfolio nominal returns. Across all portfolio structures and scenarios, there are
significant deviations from the baseline projections during the first year of the economic shocks, applied over
a five-year period starting in 2016 Q1.
26. Unhedgeable risk24
22
In the No Mitigation scenario, the Aggressive portfolio performs the worst , recording a maximum loss of
negative 45% and registering a permanent loss with returns not being restored to baseline projection
levels . This is consistent with economic theory on climate change which also calculates a loss into
perpetuity. What is different in this result is that these losses are real and generate an immediate impact
on the balance sheets of companies rather than represent ing a hypothetical future value discounted into
present day values .
On the contrary, the Aggressive portfolio recovers relatively quickly despite suffering the largest loss in
the Two Degrees scenario , and its total returns are above and beyond the baseline projection levels by
the end of the modelling pe riod. This trend is consistent with asset class performance, where economic
shocks have the largest impact on equities, resulting in the Aggressive portfolio to react the most as it
has the lar gest equity allocation weights.
Regardless of the sentiment sc enarios studied, the High Fixed Income portfolio will be least at risk of any
financial market disruption. However, this portfolio also experiences low performance and small overall
gains.
1.21 Impact on returns – by asset class and geography
Figure 12: Comparison of equity performance by geography in nominal % change across sentiment scenarios
Figure 12 shows market impacts on equity perf ormance by geography , primarily result ing from the
degree of vulnerability of each country’s economic fundamentals and responses to the applied shocks . In
the Two Degrees scenario, the US (W5000), UK (FTSE), Europe (DAX) and Brazil (BVSP) recover
quickly from losses to generate positive returns by the second year. However, a cross both scenarios, the
Chinese (SSE) stock index is the most negatively impacted in the short run – and does not recover after
-80%
-60%
-40%
-20%
0%
20%
40%
60%
80%
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2016 2017 2018
EquityValue
(Nominal,%changeQ-on-Q)
Two Degrees
-80%
-60%
-40%
-20%
0%
20%
40%
60%
80%
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2016 2017 2018
EquityValue
(Nominal,%changeQ-on-Q)
No Mitigation
In the No Mitigation scenario, the Aggressive portfolio performs worst, recording a maximum loss of negative
45 per cent and registering a permanent loss with returns not being restored to baseline projection levels.
This is consistent with economic theory on climate change, which also projects a loss into perpetuity. What is
different in this result is that these losses are real and generate an immediate impact on the balance sheets of
companies rather than representing a hypothetical future value discounted into present day values.
On the contrary, the Aggressive portfolio recovers relatively quickly, despite suffering the largest loss in the
Two Degrees scenario, and its total returns are above and beyond the baseline projection levels by the end of
the modelling period. This trend is consistent with expectations regarding asset class performance: economic
shocks have the largest impact on equities, causing the Aggressive portfolio to react the most, as it has the
largest equity allocation weights.
Regardless of the sentiment scenario studied, the High Fixed Income portfolio will be least at risk of any financial
market disruption. However, this portfolio also experiences low performance and small overall gains.
4.10 Impact on returns – by asset class and geography
Figure 12 shows market impacts on equity performance by geography, primarily resulting from the degree
of vulnerability of each country’s economic fundamentals and responses to the applied shocks. In the Two
Degrees scenario, the US (W5000), UK (FTSE), Europe (DAX) and Brazil (BVSP) recover quickly from losses to
generate positive returns by the second year. However, across both scenarios, the Chinese (SSE) stock index is
the most negatively impacted in the short run – and does not recover after three years – whereas the Japanese
(N225) stock index is the most heavily affected on average (and also in the long run).
Figure 13 shows the market impacts on fixed income performance by geography. Overall, the negative impact
on fixed income is not as significant as that experienced by equities – with the largest negative fixed income
loss recorded at 36 per cent, as compared to almost twice as much (68 per cent) for equities. This suggests that
equities are more volatile than fixed income assets in these scenario analyses, making the aggressive portfolio
react the most across the sentiment scenarios.
Figure 12: Comparison of equity performance by geography in nominal per cent change across sentiment
scenarios
27. Unhedgeable risk 25
23
three years – whereas the Japanese (N225) stock index is the most heavily affected on average (and
also in the long run) .
Figure 13: Comparison of fixed income performance by geography in nominal % change across sentiment scenarios
Figure 13 shows the market impacts on fixed income performance by geography. Overall, the negative
impact on fixed income is not as significant as equities – with the largest negative fixed income loss
recorded at 36%, as compared to almost twice as much ( 68% ) for equities. This suggests that equities
are more volatile than fixed income assets in these scenario analyses, making the aggressive portfolio
react the most across the sentiment scenarios.
1.22 Impact on returns – by industry sector
The impact on equity returns by industry is calculated based on the respective sectors’ betas, which
represent the sectors’ volatility to the respective market indices. Both levered and unlevered beta values
are compil ed annually for each industry and across countries by Professor Aswath Damodaran from the
Stern School of Business at New York University . These are then retrieved online
9
and analysed to better
model the sectoral impacts.
Beta is a measure of a sector’s volatility, or unsystematic risk, in comparison to the market as a whole. In
this study, we explored the impacts on industry sectors due to climate risks through shocks applied to the
beta values. The region -specific climate damage functions obtained in S ection 5Appendix D are
translated into shocks to the industry betas accordingly, so as to vary their volatility consistent with each
scenario .
For example, in the No Mitigation scenario , agricultural productivity falls significantly due to global
warming acceleration, leaving the agriculture sector in general subjected to greater climate risks and
hence volatility, resulting in significantly larger beta values. These shocks then propagate through
9
Damodaran , A. (2015) Current dataset for levered and unlevered betas by industry. [Online] Availa ble on:
http://people.stern.nyu.edu/adamodar/New_Home_Page/datacurrent.html (Assessed 24 Feb 2015)
-40%
-20%
0%
20%
40%
60%
80%
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2016 2017 2018
FixedIncomeTotalReturns
(Nominal,%changeQ-on-Q)
Two Degrees
-40%
-20%
0%
20%
40%
60%
80%
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2016 2017 2018
FixedIncomeTotalReturns
(Nominal,%changeQ-on-Q)
No Mitigation
Figure 13: Comparison of fixed income performance by geography in nominal per cent change across sentiment
scenarios
4.11 Impact on returns – by industry sector
The impact on equity returns by industry is calculated based on the respective sectors’ betas, which represent
volatility compared to the respective market indices. Both levered and unlevered beta values are compiled
annually for each industry and across countries by Professor Aswath Damodaran from the Stern School of
Business at New York University. These are then retrieved online8
and analysed to better model sectoral
impacts.
Beta is a measure of a sector’s volatility, or unsystematic risk, in comparison to the market as a whole. In this
study, we explored the impacts on industry sectors due to climate risks through shocks applied to the beta
values. The region-specific climate damage functions, obtained in Section 5 Appendix D, are translated into
shocks to the industry betas accordingly, so as to vary their volatility consistent with each scenario.
For example, in the No Mitigation scenario, agricultural productivity falls significantly due to global warming
acceleration, leaving the agriculture sector in general subjected to greater climate risks and hence volatility,
resulting in significantly larger beta values. These shocks then propagate through sectoral equity performances
and are used to provide clarity and comparison across countries and sectors, as well as between the Two
Degrees and No Mitigation.
Finally, within each market, the equity performance by industry (Table 8 and Table 9) is measured based on the
notional Value-at-Risk (VaR), defined in this study as the drop in performance in the worst impacted quarter
over the five-year modelling period (i.e. the worst one of twenty quarters, hence the notional 5th
percentile).
The top three worst performing sectors are the same in both scenario variants, which shows that systematic
effects (i.e. economy-wide effects) dominate over the short term (Table 8). In developed economies, the worst
performing sector is Real Estate, closely followed by Basic Material, Construction and Industrial Manufacturing.
In emerging economies, the worst performing sectors are Energy/Oil and Gas, Consumer Services and
Agriculture. Theoretically, it is possible to hedge risk by switching investments from the worst-performing
sectors to the better performing ones, e.g. Real Estate to Transport or Agriculture in developed markets in the
No Mitigation Scenario (Table 9).
8 Damodaran, A. (2015) Current dataset for levered and unlevered betas by industry. [Online] Available on: http://people.stern.nyu.edu/adamodar/New_Home_Page/
datacurrent.html (Assessed 24 Feb 2015)
28. Unhedgeable risk26
Table 8: Worst performing industry sectors across both developed and emerging economies
The box plots presented in Figures 14 to 17 show the distribution of returns for each sector. For example, in
Figure 19, the “box and whiskers” representing Real Estate is built from the 20 quarterly return figures for the
Real Estate sector in the No Mitigation Scenario, hence the top and bottom of the range are the best and worst
quarterly return figures, respectively, over five years. We define the bottom of the range as “notional VaR”.
Table 9: Best performing industry sectors across both developed and emerging economies
24
sectors, as well as between the sentiment scenarios, Two Degrees and No Mitigation.
Finally, within each market, the equi ty performance by industry ( Table 8 and Table 9) is measured based
on the notional Value -at-Risk (VaR), defined in this study as the drop in performance in the worst
impacted quarter over the five -year modelling period (i.e. the worst one of twenty quarters , hence the
notional 5
th
percentile).
Table 8: Worst performing industry sectors across both developed and emerging economies
Sentiment Scenarios No Mitigation Two Degrees
Countries Developed Emerging Developed Emerging
Worstperformingsectors
acrosscountries
Real Estate -36% -38% -19% -24%
Basic Material -26% -36% -14% -23%
Construction -26% -31% -14% -20%
Energy / Oil & Gas -24% -77% -13% -51%
Consumer Services -20% -44% -11% -30%
Agriculture -17% -40% -9% -25%
The top three worst performing sectors are the same in both scenario variants , which shows that
systematic effects dominate over the short term (i.e. economy -wide effects) (Table 8). In the developed
economies, t he worst performing sect or is Real Estate closely followed by Basic Material , Construction
and Industrial Manufacturing. In the emerging economies, the worst performing sectors are Energy/Oil
and Gas, Consumer Services and Agriculture. Theoretically, it is possible to hedge risk by switching
investments from the worst -performing sectors to the better performing ones, e.g. Real Estate to
Transport or Agriculture in developed markets in the No Mitigation Scenario (Table 9).
Table 9: Best performing industry sectors across both developed and emerging economies
Sentiment Scenarios No Mitigation Two Degrees
Countries Developed Emerging Developed Emerging
Bestperformingsectors
acrosscountries
Transport -17% -27% -9% -17%
Agriculture -17% -40% -9% -25%
Consumer Retail -20% -26% -11% -18%
Health Care -21% -23% -11% -15%
Industrial/Manufacturing -25% -26% -13% -16%
Technologies -21% -27% -11% -16%
24
notional 5 percentile).
Table 8: Worst performing industry sectors across both developed and emerging economies
Sentiment Scenarios No Mitigation Two Degrees
Countries Developed Emerging Developed EmergingWorstperformingsectors
acrosscountries
Real Estate -36% -38% -19% -24%
Basic Material -26% -36% -14% -23%
Construction -26% -31% -14% -20%
Energy / Oil & Gas -24% -77% -13% -51%
Consumer Services -20% -44% -11% -30%
Agriculture -17% -40% -9% -25%
The top three worst performing sectors are the same in both scenario variants , which shows that
systematic effects dominate over the short term (i.e. economy -wide effects) (Table 8). In the developed
economies, t he worst performing sect or is Real Estate closely followed by Basic Material , Construction
and Industrial Manufacturing. In the emerging economies, the worst performing sectors are Energy/Oil
and Gas, Consumer Services and Agriculture. Theoretically, it is possible to hedge risk by switching
investments from the worst -performing sectors to the better performing ones, e.g. Real Estate to
Transport or Agriculture in developed markets in the No Mitigation Scenario (Table 9).
Table 9: Best performing industry sectors across both developed and emerging economies
Sentiment Scenarios No Mitigation Two Degrees
Countries Developed Emerging Developed Emerging
Bestperformingsectors
acrosscountries
Transport -17% -27% -9% -17%
Agriculture -17% -40% -9% -25%
Consumer Retail -20% -26% -11% -18%
Health Care -21% -23% -11% -15%
Industrial/Manufacturing -25% -26% -13% -16%
Technologies -21% -27% -11% -16%
31. Unhedgeable risk 29
This implies that just slightly less than half of the returns impacted due to climate change can be hedged
through cross-industry and regional diversification.
Table 10: Summary of portfolio performance measured by the 5 per cent VaR by structure and scenario,
nominal per cent
Table 11: Summary of portfolio performance (long-term impact after 5 years) by structure and scenario,
nominal per cent
Interestingly, we find a similar magnitude for the hedging potential of different portfolio allocations. Table
10 and Table 11 consider the notional VaR and long-term impact by portfolio structure and scenario. From
this we can infer that, under No Mitigation, 49 per cent (=(45-23)/45) can be hedged by shifting from more
equity-loaded portfolios (Aggressive) to a fixed income-heavy portfolio (High Fixed Income). Contrasting this
to the portfolio’s longer-run performance, portfolio managers can use the hedging potential across portfolio
structures to maximise gains up to 28 per cent by shifting from a high fixed income investment to an equity-
loaded structure (Aggressive) under Two Degrees.
An investment manager wishing to hedge climate risks for both Two Degrees and No Mitigation scenarios is
advised to adopt the High Fixed Income portfolio containing assets from developed markets. Even though
long-term returns are low in this case, downside losses are minimised. In the event of a Two Degrees scenario,
there is little opportunity to hedge downside climate risk through portfolio construction. In the event of a Two
Degree scenario an Aggressive portfolio offers the best positive returns over the long term. In the event of a No
Mitigation scenario, the High Fixed Income scenario offers both the best protection against downside risk and
the best long-term performance.
27
enable investors to shield themselves from different scenarios of climate risk .
By analys ing sectors in developed and emerging markets ( in Figure 14 to Figure 17), we can reveal the
hedging potential of cross -industry diversification and investment in sectors with low climate risk . We find
that under No Mitigation, the wo rst-case scenario, it is possible to cut the maximal loss potential by up to
47% by shifting from Real Estate (in developed markets) and Energy/ Oil & Gas (in emerging markets)
towards Transport (in developed markets) and Health Care/ Pharma (in emerging markets). This implies
that just slightly less than half of the returns impact ed due to climate change can be hedged through
cross -industry and regional diversification.
Table 10: Summary of portfolio performance measured by the 5% VaR by structure and scenario, nominal %
Portfolio Structure Baseline Two Degrees No Mitigation
High Fixed Income 0 -10% -23%
Conservative 1% -11% -36%
Balanced 1% -11% -40%
Aggressive 1% -11% -45%
!!
Table 11: Summary of portfolio performance (long -term impact after 5 years)
by structure and scenario, nominal %
Portfolio Structure Baseline Two Degrees No Mitigation
High Fixed Income 4% -3% -4%
Conservative 12% 9% -26%
Balanced 16% 17% -30%
Aggressive 21% 25% -45%
Interestingly, we find a similar magnitude for the hedging potential of different portfolio allocations. Table
10 and Table 11 consider the notional VaR and long -term impact by portfolio structure and scenario.
From this we can infer that, under No Mitigation, 49% (=(45-23)/45) can be hedged by shifting from more
equity-loaded portfolios (Aggressive) to a fixed income -heavy portfolio (High Fixed Income). Contrasting
this to the portfolio’s longer-run performance , portfolio managers can use the hedging potential across
portfolio structures to maximise gains up to 28% by shifting from a high fixed income investment to an
equity-loaded structure (Aggressive) under Two Degrees.
An investment manager wishing to hedge climate risks for both Two Degrees and No Mitigation scenarios
is advised to adopt the High Fixed Income portfolio containing assets from developed markets . E ven
though long-term returns are low in this case , downside losses are minimised . In the event of a Two
Degrees scenario, there is little opportunity to hedge downside climate risk through portfolio construction.
In this scenario an Aggressive portfolio offers the best positive returns over the long term. In the event of
a No Mitigation scenario , the High Fixed Income scenario offers both the best prote ction against
downside risk and the best long-term performance.
32. Unhedgeable risk30
5 Conclusion
This study has quantified the implications of climate risk on investment portfolios over the short term. As far as
we are aware, this is the first time that research addressing this question has been attempted. Previous studies
have analysed the direct physical effects of climate change over the long term, typically in the latter half of this
century, when the effects of climate change will have already started to have major impacts. These studies then
discount the future impacts of climate change and provide an estimate in net present value terms. However,
financial markets could be affected much sooner by market sentiment, which may change as new information
comes to light about the effects of climate change.
The innovative approach adopted in this research allows us to simulate ‘what-if’ scenarios relevant for the next
five years. The effects of climate change on markets will be driven by the projections of likely future impacts,
for instance from new technology, changing regulation, indirect climate change impacts and shifting market
sentiment. The scenarios we developed are coherent, quantifiable narratives. Although highly unlikely, they
still are plausible for describing how expectations about future climate trajectories may have an impact on
economic and financial markets over the next five years. This study therefore quantifies the potential financial
impacts of a shift in market sentiment, driven by significant changes in investor and consumer beliefs about the
future effects of climate change.
We find that, even in the short term, climate risks pose a significant threat to investment portfolio performance.
In the worst-case scenario, a global recession occurs during the first three quarters of the shock and the global
economy never recovers, losing an estimated 16 per cent of economic output by 2050. In the alternative
scenario, action to limit warming to within 2°C will have negative short-term costs, lowering economic output
for over a decade when compared to baseline. However, the long-term benefits make the transition worthwhile,
increasing aggregate output to 2050 by up to 4.5 per cent above baseline projection levels. Investors are
therefore encouraged to take a long-term perspective when considering climate risk in investment portfolios.
Nevertheless, investors cannot entirely shield themselves from the exposure to climate change. Although we
have shown that 47 per cent and 49 per cent of impacts due to climate change can be hedged through cross-
industry and portfolio construction, these two “halves” are not cumulative such that no one strategy is able to
offer more than 50 per cent coverage of “hedgeable” risk. Climate risks therefore will remain an aggregate risk
driver that requires system-wide action to mitigate its economy-wide effects.
6 References
van Aarle, B. and Kappler, M., 2012. Economic sentiment shocks and fluctuations in economic activity in the
euro area and the USA. Intereconomics, 47(1), p.44–51. DOI:10.1007/s10272-012-0405-z.
Ackerman, F. and Stanton, E., 2006. Climate change: The costs of inaction., Global Development and Environment
Institute, Tufts University. http://cursa.ihmc.us/rid=1H0VSPQK1-CHJ1Z7- JJP/$Tper cent20Costper cent20ofper
cent20Climateper cent20Inactionper cent201006.pdf [Accessed September 15, 2015].
Alexander, L., Allen, M., Intergovernmental Panel on Climate Change and Working Group I, 2013. Climate
change 2013 the physical science basis: summary for policymakers : Working Group I contribution to the IPCC
fifth assessment report., Geneva: WMO, IPCC Secretariat. http://www.ipcc.ch/report/ar5/wg1/#.Up43xtJDtYo
[Accessed July 15, 2015].
Arias, A., 2014. Sentiment Shocks as Drivers of Business Cycles. , p.1–36.
Association of British Insurers, 2005. Financial Risks of Climate Change. Technical report http://insurance.lbl.
gov/documents/abi-climate.pdf [Accessed July 21, 2015].
